Abstract
Side comments and conversations in focus groups can pose challenges for facilitators. Rather than seeing side comments as problematic behavior or “failed” data, we argue that they can add to and deepen analyses. Drawing on focus group data with grade nine students from a study on early work, in this methodological paper we discuss three patterns. First, side comments have highlighted where participants required clarification, and illustrated their views and questions about the research process. Second, side comments added new data to our analysis, including personal reflections, connections to others’ comments, and information about participants’ uncertainties about the research topics. Third, these comments offered insight into peer relations and dynamics, including participants’ reflections on age, and how they deployed gender relations in their discussions. Provided that their use fits within established ethical protocols, we argue that there is a place for attention to side comments, especially in focus group research with young people where adult-teen hierarchies and peer dynamics might lead young people to engage more with peers than directly respond to researchers’ questions.
Introduction
In the middle of a focus group on teens’ thoughts about doing paid employment, Adrianna and Austin started a side conversation that was picked up by a digital recorder but was not part of the main focus group conversation. They were arguing about whether another student in the class had actually worked as a lifeguard. Austin was doubtful and suggested that the girl was lying. Adrianna said that she had inside information to suggest that it was the truth. The conversation then shifted onto whether Austin could swim and how fast. This was one of many engaging side conversations that happened during some of our focus groups that prompted us to think about the boundaries of focus group data, and to engage with rich, insightful data that might have otherwise been overlooked. Despite their disengagement from the main focus group conversation, Adrianna's and Austin's side conversation was not entirely off track from the topic of the focus groups, which was early work. Their lively interaction is exemplar of the ways that participants in our focus groups moved in and out of the main focus group conversation, and between topics that were closer or farther away from our initial research agenda. We aim to provoke thinking about the boundaries of data by showcasing the value in engaging with data from such side conversations and discussing their possibilities and challenges for focus group methodology.
Focus groups with young people can range from orderly, serious engagements, to boisterous conversations (Allen, 2005; McGarry, 2016; Raby, 2010). Many things can go awry in focus groups, including focus groups with children and youth, such as participants’ silliness, exaggeration and bravado, dominant and dominating voices, crosstalk, and jockeying for participation (e.g., Allen, 2005; Hyde et al., 2005; Raby, 2010). We do not seek to evaluate such focus groups as successes or failures, however. Rather, we aim to add a perspective to focus group methodology centered on the insights that can be gained when researchers facilitate in ways that open possibilities for multiple forms of participation—including side conversations—and engage in analyses which grapple with data which may appear random or off-topic. Thus, we position side conversations in focus groups as a particular and insightful instance of interaction, including aforementioned dimensions like silliness, mocking, and bravado.
In this methodological paper, we reflect on eight focus groups conducted with 13- and 14-year-olds about their experiences and views around very first jobs, including things like babysitting, snow-shovelling, and working in low level service jobs. These eight focus groups all occurred in a school cafeteria one December day in 2018. Some of these groups were quite active, with many main and side conversations happening at the same time, leading us to reflect on what we can learn from such comments in focus groups. While side comments and conversations led to interruptions in the focus groups and sometimes made transcribing difficult, we asked what they could tell us, even when they might seem off track. These side conversations have provided insight into how our participants engaged with the research process itself, including where they might have needed clarification about expectations. They also provided new data, including their personal reflections on the topic of conversation and how participants engaged with each other around our questions. Finally, these side comments fostered insight into peer dynamics, including those related to age, gender, and work. Reflecting on the boundaries of data collection and the ethics of attending to side comments, we argue that the parameters of our specific project and our care in sharing the data support the ethical use of these side conversations as data, although this might not be the case in other projects. We argue that side conversations, when collected transparently and shared ethically, can provide research insights, especially in research with young people, as adult-teen hierarchies and peer dynamics in focus groups might facilitate engagement with peers, rather than responding directly to adult researchers’ questions.
Literature review: facilitation, dynamics, and analysis in focus groups with young people
Focus groups are used quite frequently when conducting research with young people. We first review some of the literature on focus groups with young people, with a particular emphasis on how they can shift power relations with adults, and then on how focus groups provide us with ways to learn about how people make meaning in groups. We also reflect on how to best work with what might sometimes feel like “failed” or overly “unruly” focus group data.
One of the strengths of focus groups is the interaction that occurs within the research encounter, including between participants themselves (Halkier, 2010; Kitzinger, 1994; Morgan, 2010). Although not suitable for all projects, focus groups and the interactional dynamics within them, can provide insight into how people make meaning in groups (Hollander, 2004; Warr, 2005), and how participants react to and engage with a research topic, as well as the specific research dynamic (Katainen and Heikkilä, 2020). For instance, Hydén and Bülow (2003) argue that in focus groups, participants establish a common ground, and then either add to or divert from this common ground. Informed by Goffman, Halkier (2010) positions focus groups as a particular form of social interaction and argues that much of the conversation within focus groups could be best understood through examining this method's relation to social norms and discourses. Attention to reproductions and rewritings of social norms in focus groups may thus reveal participants’ relations to others, and their identity making processes (Halkier, 2010). For Kitzinger (1994), these interactions are shaped by the characteristics and dynamics of the group so attention should not only focus on what is said, but also on how the focus groups’ dynamics shape conversation (Kitzinger, 1994).
Participant-researcher hierarchies position researchers as elevated, knowledgeable experts in research contexts (e.g., Hoffmann, 2007; Smithson, 2000). In research with young people, these hierarchies are further complicated by age differences and related discourses of childhood innocence and adult competency, which locate adults as competent and rational, and young people as naïve, vulnerable, and lacking experience and expertise (Katainen and Heikkilä, 2020; Raby, 2010; Spyrou 2018). Focus groups can disrupt these researcher-participant hierarchies, however, through participants’ greater numbers and shared peer connections (Raby, 2010), allowing them to challenge facilitators’ assumptions about their experiences, views, and interests (Allen, 2005; Katainen and Heikkilä, 2020; Raby, 2010). Focus groups also increase opportunities for participants to produce meaning together (Hollander, 2004), as well as challenge other participants’ ideas and researchers’ agendas (Jowett and O’Toole, 2006). We must thus recognize power as complex and relational within focus group research.
Facilitating focus groups, including with teens, requires balancing between directing the conversation and letting the conversation flow (Kitzinger, 1994; McGarry, 2016). Drawing on focus groups with young men in a project on masculinity, Allen (2005) argues that researchers should not try to control or minimize potential identity work that may occur between participants in a focus group, including possible masculine bravado. Rather than seeing potential identity work as necessitating intervention by facilitators to deter possible negative effects on the data, potential identity work in focus groups can offer insight into peer dynamics, such as how gendered identities are negotiated (Allen, 2005). Echoing Allen's (2005) call for reflexive rather than interventionist facilitation, McGarry (2016) positions a more flexible, laid-back, and reflexive engagement by facilitators as aligned with seeing young people as dynamic and knowing.
The interaction dynamics and patterns within focus groups are also shaped by other social forces, including power dynamics, within and beyond the focus group (Hollander, 2004; Katainen and Heikkilä, 2020; Vitus, 2008) and connected to factors such as gender, age, race, and social class (Allen, 2005; Halkier, 2010; Nairn et al., 2005; Kitzinger, 1994). In focus group research with young people in Finland, Katainen and Heikkilä (2020) found that participants’ social class backgrounds, including their social and cultural capital, shaped how they responded to the research process and their interactions. They found that active/engaged participants were those most likely to hold social and cultural capital (Katainen and Heikkilä, 2020). Smithson (2000) similarly notes that some participants have more dominant voices and may reproduce social discourses and norms in ways that limit possibilities for dissent in a focus group. Smithson (2000) cautions that when an individual dissents from the rest of the focus group they may be constructed as an “Other,” a pattern which can be further complicated by intersections of race, gender, class, and age. Importantly, focus group interaction patterns are also shaped by the moderator's social positioning, adding nuance to the complexities of critical, reflexive facilitation (Smithson, 2000).
Given the relevance of group dynamics, and the emerging and often unruly interactions that occur within focus groups, it is possible for discussions to veer away from researchers’ plans by becoming chaotic and overwhelming to facilitate, transcribe and analyze (Warr, 2005). Focus groups can also be challenging when participants are very quiet, sarcastic, or inclined to use humor to deflect questions. Rather than seeing these as indications of a “failed” interview or focus group (Jowett and Toole, 2006) various authors note that these challenges can provide important insights, including around resistance to the research process and agenda. For instance, reflecting on an unusually quiet, awkward focus group, Nairn et al. (2005) discuss how data is not only provided through speech but also things like researcher and participant embodiment of certain identity categories and hierarchies; the context of the focus group (e.g., a school classroom); pauses and mumbling; and silences and laughter, which can reflect discomfort but also a “safe” refusal or resistance. By positioning resistance as emerging in context, and in relation to the people, place, and objects (Tuck and Yang, 2013), we can see how young people's style of participation, including how they respond to questions and the research space, can provide insight into flows of power, agency, and resistance in focus groups. Jacobsson and Åkerström (2012) similarly discuss how a participant might have an agenda which counters the intent of an interview, but again, this can be considered a dynamic to learn from rather than an interview failure. Such “counter-talk” strategies may allow participants to “maintain worth and dignity” (Heikkilä and Katainen, 2021: 1032), manage anxiety, and to shift (or retrench) class, race, gender, and age-based power relations.
Echoing Jowett and Toole (2006), we resist a framing of an “unruly” or “off track” focus group as a “failed” focus group, instead we seek to better understand how the interaction that occurs within focus groups, specifically in the side conversations between participants adjacent to the main research discussion, provide us with insight into our research context and participants' thoughts and questions about the research process, add depth to the data, and illuminate peer dynamics.
Specific ethical dilemmas also arise in focus groups with children and youth, that are related in part to the researcher-participant hierarchy (Morgan et al., 2002). For example, Morgan et al. (2002) advise facilitating focus groups with children in locations that might lead to more casual engagement, for example, a community center room, rather than a school, which may lead participants to see researchers as teachers. Importantly, researchers should reassure young participants that they are not being evaluated, that there are no “right” answers, and that their ideas are valuable (Gibson, 2007). Use of scenarios and role-playing activities are also suggested in such focus groups because they foster discussion of sensitive or difficult topics without requiring participants to draw explicitly on personal experience (Morgan et al., 2002). Researchers can also encourage participation, especially among shyer participants, without intervening too much into peer dynamics (Hyde et al., 2005; Morgan et al., 2002). Group confidentiality can also be important but potentially challenging when young participants belong to the same peer group, school, or community organization (Gibson, 2007; Hyde et al., 2005).
Researchers have argued that most of the literature on focus groups is concentrated on facilitation styles and their implications, with much less on the analysis of focus group data, including the complex interactions that occur in focus groups (Belzile and Ӧberg, 2012; Duggleby, 2005). Importantly, assumptions about the usefulness of the form and content of interaction between participants shape how researchers transcribe data, and its subsequent analysis (Belzile and Ӧberg, 2012; Duggleby, 2005; Myers and Lampropoulou, 2016). Researchers’ decisions on what to include about interactions, such as laughter or non-seriousness, are connected to whether such interactional details are considered important (Myers and Lampropoulou, 2016). When included in transcripts and writing, Duggleby (2005) argues that details about the interaction patterns can help contextualize the research as a social encounter. Farnsworth and Boon (2010: 620), explain that “there is a level of interaction and dynamics that runs entirely parallel to the information gathering process of the focus groups,” but that often these verbal and non-verbal dynamics are problematically left out of transcripts and analysis.
Attention to the interaction dynamics in focus groups might also help us to move beyond discussions of focus groups as having either an individual or group unit of analysis (see Morgan, 1995; Kidd and Parshall, 2000). Attention to consensus and difference in focus groups (Morgan, 1995), might encourage an analytical turn towards interactions and dynamics, and away from generalizations about the “group” (Kitzinger, 1994). Attention to what the interaction does in focus groups (Belzile and Ӧberg, 2012) provokes thinking about interaction and group dynamics as co-constituting the focus group flow and topic (Warr, 2005). Thus, attending to interaction, and interactional dynamics, including side conversations, can offer pedagogical potential and reflection on methodology and method.
When transcribed, details about the focus group environment, as well as interaction between participants, such as laughs, silences, interruptions, challenges, and side conversations, although “unruly,” add deep context and vitality to the narrative of the focus group and analyses (Duggleby, 2005; Vicsek, 2007). For Morgan (2010) such interaction does not produce the data, it is the data. We add to this discussion, articulating the potential that can be found in the chaotic pandemonium of some of our focus groups with teens, specifically the side conversations participants engaged in alongside the main research discussion. In all, we position unruly data from side conversations as valuable for appreciating the knowledge and experience that young people have, providing added insight into the topic at hand, learning how the research process is being experienced, and illustrating how meaning is produced through interaction.
Our focus groups on early work
Our data emerges from a broader project on young teens’ views on early work and early work experiences. In Ontario, many teens first start working part-time for pay when they are fifteen years old, although some start working when they are younger. In this paper, we draw on eight focus groups with 65 grade nine students (ages 13 and 14), 30 boys and 35 girls, that were conducted during one day at a publicly funded, Catholic high school in a fairly economically secure area of a small city in Ontario, Canada. Forty participants identified as white, three as Latinx, four as Black, five as Asian, six as mixed-race, and seven did not self-identify in terms of race or ethnicity. Participants were from diverse class backgrounds.
We facilitated semi-structured focus groups because we were interested in how young people talk between themselves about early work (Halkier, 2010; Kitzinger, 1994; Morgan, 2010), including their experiences of early work, their views on children working, and how they would manage difficult workplace scenarios. Our questions were informed by our interviews with young workers in an earlier project. We conducted six of the focus groups in a noisy high school cafeteria and two in a classroom, with groups of between eight and ten students. Focus group sizes were determined by the number of students in a class who opted to participate during each school period, and the large size of the groups likely contributed to the prevalence of side comments. Each focus group lasted about 50 minutes. The students knew each other, as they were all in classes together, and some of them were friends. The students who participated had parental consent and provided their own assent.
In their groups, participants first completed short, individual surveys about their early work experiences. We then explained the focus group format and asked them about their ideas on early work and about any previous early work experiences. Next, we presented several workplace scenarios or vignettes that presented difficult workplace situations (e.g., Church and Ekberg, 2013; O’Reilly and Parker, 2014; Wong et al., 2018) and asked them what they might do if they were in the situation being described. Each focus group was led by one facilitator. As facilitators, we asked the questions, guided the students to stay on track, asked them to take turns speaking, and aimed to move the focus groups through the questions within the time available. Given the anticipated size of the groups, we used two recording devices in each focus group and had a research assistant doing on-site transcribing. Students also had name cards in front of them. We invited them to choose their own pseudonyms for our reporting of the data, which we use here.
The focus groups were all transcribed, using the transcript produced by the on-site transcriber as a starting point, and then filling in gaps by listening to the two recording devices. We then coded the transcripts, first conducting open coding and then moving towards more abstract coding with NVIVO and then QUIRKOS qualitative analysis programs. Our coding included attention to focus group dynamics, as we were particularly interested in the many layers of conversations, interruptions and side comments that were happening during some of the groups.
As we listened (and relistened) to the audio recordings while transcribing, we were able to identify multiple simultaneous conversations throughout the recordings. We categorized the main research discussion as that which was organized around the activities and questions guided by the facilitator. Side conversations were those which occurred simultaneous to the main research discussion, but between small groups of participants without talking directly to or with the facilitator and participants in the main research discussion. Sometimes these side conversations were related to the main research discussion, as we discuss in our analysis, and other times these side conversations were about the research process and topic prompts, disagreements between participants, and silliness. These side conversations were not neatly enclosed and contained, instead, participants often moved between side conversations and the main research discussion.
During our engagement with the side comments, we reflected on the context of the focus groups. The focus groups were conducted right before the winter break and students who participated left their regular classes to meet with us, so there was an atmosphere of restlessness and excitement among the participants. For most of the focus groups, the students joined us in the school cafeteria, an open room with long tables and benches, bulletin boards with posters advertising school events, and a galley-style kitchen with friendly cafeteria staff. While two focus groups happened in classrooms with a small circle of desks pushed together, the cafeteria focus groups had a far greater number of side conversations. A school cafeteria is a space where students usually enjoy a break from class, socialize, and connect with peers. Indeed, a cafeteria is a space in a school where students can be energetic, loud, and chatty—behaviors which tend to be discouraged in a classroom setting. School cafeterias can also be spaces which are stressful, riddled with peer cliques, and sites of peer conflict. Hopkins (2007: 534) reminds us to “analyz[e] the ways in which particular locations, places and time of day … influence focus group discussion.” Thus, the time of year, the physical space of the school cafeteria, the organization of tables, and how the cafeteria was usually occupied, likely affected the ways participants engaged with the focus groups. Some participants may have felt particularly comfortable engaging in crosstalk and side conversations with their peers in this space, as that way of engaging might be “natural” during their lunch break. A different space might have felt less comfortable for side conversations with peers, which seemed to be the case with the two focus groups conducted in a classroom.
Before engaging with the focus group data, we must also reflect on our positionality as researchers, our facilitation styles, and questions around ethics. Both authors facilitated the focus groups, Rebecca as a white, middle-aged professor, experienced facilitator and the parent of a teen, and Lindsay as a white, MA student and young adult who was new to facilitating. Rebecca tended towards a looser facilitation style, with a lot of room for the kind of side-conversations that we are examining in this paper. Her focus groups also seemed more likely than Lindsay's to be mixed-gender groups and to have boisterous friends in the focus groups together. Lindsay tended towards a more formal facilitation style and her focus groups also seemed to include more girls, and quieter students who needed to be encouraged to speak. We have wondered, however, whether Rebecca's groups were more conducive towards side conversations partly because of how we were perceived by the participants. It may be that because Rebecca was seen as more separate from the participants’ age and peer contexts, it allowed, invited, or even pushed participants towards greater peer-to-peer engagement as the participants sought to carve out a distinct space for themselves and to ensure their control over the group dynamics. In contrast, Lindsay may have been seen as more of a peer.
Ethics
This project received ethics clearance from two universities and was approved by a school board. Nonetheless, a reader might wonder whether it is ethically appropriate to engage with participants’ side comments in a focus group. Did participants intend their side comments to be part of the data? What kinds of side comments are ethical to include and what kinds are not? Does attention to such side comments compromise participants’ confidentiality?
Researchers have sometimes been concerned that when the researcher is familiar with participants, participants might speak more freely than intended (Thorne, 1980). In a focus group where participants know each other, they might similarly let their guard down. We do not feel that this was the case in our project, however. Each focus group was conducted in a public space, with classmates who were not all friends, all crowded close together at a table. Participants were sharing the table with the main facilitator on one end and the transcriber on the other. There was also a recording device on each side of the table. At the beginning of the focus group, we reviewed the ways the recorders and transcription worked. For these reasons we feel that the side comments that were picked up by the recorders were acceptable for us to use, particularly as the kinds of side comments that were picked up by our digital recorders were not deeply personal. Finally, in sharing the data, we have covered standard procedures to maintain confidentiality by avoiding identifying information about any of our participants. In some cases, we were unable to even identify who was speaking in the side comments, thus they are anonymous.
It may still be the case that some of the participants did not intend their comments to be used as part of the data. Comments were frequently made in an off-hand manner, for instance, and out of earshot of the main focus group facilitator, as part of peer-to-peer banter. However, some of the side comments were made directly into the digital recorder or made with reference to the recorder, suggesting that the speakers intended to be recorded. One concern about focus groups is that some participants may not feel comfortable speaking up in a group, especially if they disagree with the group or feel silenced due to hierarchies within the group (Hollander, 2004). Rather than being data that is ethically suspect, we thus see some of these side comments as participants embracing an opportunity to share their views with us without engaging with the whole group.
Analysis
We now present three patterns that emerged in our analysis of the side conversations. First, these conversations highlighted areas where participants were seeking clarification around the questions we were asking and/or were commenting on the research process itself. Second, these side conversations provided answers to some of our questions. Third, these side comments provided us with insight into peer relations and dynamics.
Pattern one: clarification and the research process
The first pattern illustrates how side conversations offered insight into where participants required clarification about the research project, as well as their engagement with the research process. For example, participants sometimes sought clarification around the work scenarios that we presented to them. The scenario that generated the liveliest conversation and debate was around deciding whether to meet a previous babysitting commitment or play in an important baseball game that was scheduled at the last minute. In one side conversation, two girls, Maisy and Mackenzie, clarified the scenario for another participant. If you said yes to babysitting first, you can’t back out. Especially not on the day of, like if you get told about the game on the day of … […] What? But it depends on when you get told about the babysitting. Like if you get told about the babysitting the day of, you can [inaudible] No, this is the scenario though: you get asked to babysit and you say “yes” and then you figure out the game. Yeah, then you get told [about the game], like the day of. Like, what are you choosing?
In this exchange we see that some of the participants were a bit confused by the scenario, and how they were trying to work together to clarify it. This clarification gets mixed in a bit with their answer to the question, suggesting that the students wanted to interpret the scenario correctly as they commented on it, and indicating that the specific structure of the scenario shaped their responses to it. Thus, through our engagement with the side conversations, we were able to reflect on how our approach as facilitators and the format of our questions and activities shaped how participants engaged (see also Katainen and Heikkilӓ, 2020; Smithson, 2000).
A second example arose around the recording equipment. Before turning on the recording devices and placing them at opposite ends of the cafeteria tables, we explained that they recorded audio only, which was also noted in the letters of consent. The equipment seemed to provoke questions, curiosity, and silliness. For instance, despite our explanations, Nate (a boy) and Berry (a girl) wondered if it was a camera or microphone. Their musings are concerning as they illustrate that participants did not always hear or understand our introduction to the research, but they are also helpful in letting us know this. Additionally, they suggest a lack of familiarity with the equipment, and potentially a self-consciousness around being recorded. In another focus group, a girl with the pseudonym of K and a boy, Thad, spent lots of time having side conversations and fiddling with the recording device (e.g., tapping on it, blowing into it, whispering into it). Sally, another participant, asked K “Can you stop making noises in this thing?” Eventually Rebecca also asked K and Thad to stop making silly noises into the recorder, explaining that it makes transcription more difficult. K got confused or just silly, wondering if someone was inside the recorder. Rebecca then explained what is meant by transcription and K whispered “I am so sorry” into the recording device. This last example may indicate a kind of resistance to the research project (Heikkilä and Katainen, 2021; Vitus, 2008) through being silly, but alternatively could illustrate how participants can be unfamiliar with research processes. It also reminds us that we cannot categorize a single focus group as a homogenous mass, as some participants were bored or distracted while others wanted to pay attention. In one group, our recording even picked up a girl whispering, “who really cares?” during a conversation on negotiating babysitting hours. Similarly, in another group, after Rebecca explained “so I have uhm, I might have a few more questions for you, but first I have an activity,” an unknown participant sighed “uh, questions,” and another responded with “let's go! [let's answer the questions!]” Here, we see one participant who seemed less interested in participating in the focus group or engaging with more questions, and one participant who was more enthusiastic. In this interaction, the more enthused participant seemed to try to encourage the other.
These moments prompt us to think about the complexity and possibilities of focus groups with teens, and to reflect on research design. Resonating with arguments made by Jowett and Toole (2006) and Katainen and Heikkilä (2020), we see these examples, once transcribed, as moments in our analysis which made us think more deeply about how participants reacted to and engaged with our research project. Thus, we agree that no focus group “fails” in an absolute sense (Jacobsson and Åkerström, 2012; Jowett and Toole, 2006), as the dynamics and interactions between participants can provoke critical reflection on focus group methods and provide information about participants’ engagement.
There are other examples where participants tried to manage other participants’ behavior and attention. In one, during a focus group, a boy, Ice, noticed a friend who was not part of the focus groups enter the cafeteria, and he shouted “hello!” to them. Berry seemed bothered by this disruption and distraction, and told Ice and others to “focus, guys!” Then Ice responded to Berry, saying “I was just saying ‘hi’ to my friend.” With Belzile and Ӧberg (2012), we can interpret these examples with attention to what the interaction does for how the conversation flows and participants’ engagements. In these cases, the interaction in the side conversations, including how participants responded to the focus group process, their inattention, and the gendered managing of others in the groups shaped how the focus group unfolded. In these examples we see how when some participants are distracted and inattentive others respond to them, sometimes trying to steer the conversation back to the main research discussion.
Pattern two: adding to data
Our second pattern focuses on how side conversations provide data about the topic at hand, including more personal reflections and participants building off others’ comments. They illustrate how depth and meaning is produced through interaction (Kitzinger, 1994; Raby, 2010), but not always in ways that are directly evident to a focus group facilitator, especially in larger and more chaotic focus groups. One example of this insight illustrated divergent viewpoints when the participants were discussing a hypothetical scenario that involved a manager asking a young worker to use a broken meat slicer. A consensus was emerging in the group where participants were saying that they would seek independent solutions or continue working with the broken machine to keep their paid work. Our recording picked up Austin muttering a firm, contrasting stance saying “yeah, just slice yourself open for 15 dollars, f*** that!” One issue with focus groups is that individual participants sometimes feel that they cannot dissent from the group (Smithson, 2000), but here we have heard this important dissent, albeit indirectly.
In another instance, while the main focus group was talking about challenges they had experienced in paid work, workplace safety came up in a side conversation between two boys, both with some work experience. Cameron advised Jake: “don’t walk in a kitchen in running shoes … cuz the floors are always greasy, it's slippery.” We learned here about some participants’ safety knowledge and how peers can and do educate each other in terms of workplace safety. In other words, as well as providing feedback about the clarity of our scenarios, such asides allowed us to learn about their views and potential strategies around navigating safety concerns at work.
In another side conversation, two girls, Maisy and Mackenzie, debated whether they would quit extracurricular commitments if necessary, to create time for work to help support their families:
What doesn’t matter? You need money for your family.
Yah, that's what the point is.
Do you go to extracurriculars?
I don’t have extracurriculars
K, well if you did. Or would you give that up to work?
You should be nice to family.
Yeah, by working to help them.
Yeah, but that doesn’t answer the question.
Yes, it does.
Well, all you said at first was “be nice.”
Shh.
Maisy and Mackenzie are engaged enough in the focus group conversation to have a side conversation about the topic that offers meaningful data on the importance of supporting family through working if necessary, and illustrates how participants make meaning together (Hollander, 2004; Warr, 2005). We also see Maisy attempting to take on the role of the facilitator, challenging Mackenzie to clarify and support her response and steering the conversation back to the initial question. Interestingly, unlike Mackenzie, Maisy had previous work experience, which may be connected to her assertiveness in this interaction (see Katainen and Heikkilä, 2020). Echoing arguments made by other scholars, in this example we see focus groups with teens as opportunities to complicate adult-teen/interviewer-interviewee hierarchies, wherein adult facilitators are positioned as knowing experts (Allen, 2005; Katainen and Heikkilä, 2020; Raby, 2010). Further, this exchange again illustrates how dissent occurs within side conversations in focus groups, in ways that might not be evident in the main research discussion (see Smithson, 2000). Maisy's and Mackenzie's interactions show us that peers may challenge each other, and through their interaction, create knowledge together (Hollander, 2004; Jowett and Toole, 2006).
As we have seen in the side conversations, particular dynamics in the focus group, including the relevance of the timing and location of the focus groups, the facilitators’ style, and the peer-to-peer dynamics, affect how participants engage with the research and each other. In the side conversations, participants worked to create their own agenda for the focus group, including how and if they engaged with the topics of the main research discussion, how they engaged with their peers and the researcher, and how long they discussed topics. These peer-to-peer interactions and dynamics offer insight into how participants respond to and push back against the research agenda, including the topics, and the presumed role of participants and facilitators (Hollander, 2004; Jowett and Toole, 2006; Katainen and Heikkilä, 2020; Raby, 2010). We argue that these details which emerge, in part from side conversations, add significantly to the depth and vitality of the data (Duggleby, 2005), and offer insight into participants’ interactions. However, such details might be non-existent or overlooked when researchers adopt a facilitation style which deters or prevents focus group engagement that might seem disruptive, disengaged or “off track” (see Allen, 2005), or when side comments are overlooked in transcription and analysis.
Pattern three: peer dynamics
Our final pattern focuses on how side comments illuminated peer dynamics, with a particular focus on gender. First, we focus on how participants attempted to influence each other around views on work. We then look more directly at how they encouraged and discouraged each other in terms of speaking up, which seemed to censure certain students over others.
Participants’ general views on work were prominent throughout the main focus group discussions but there were also side comments that pointedly illustrated peer pressure regarding whether young people should work. In one example, we asked participants for reasons why teens their age might work or not. Nate suggested that teens might work in high school because they have free time. In a side conversation, another boy, Ice, challenged Nate's idea:
What? You wanna work in your spare time?
Well, if you have already done everything and you … even relaxing.
You haven’t done everything, though. [Another participant agrees.]
This exchange seemed to be about Ice critiquing Nate for suggesting that he might work in his spare time, countering that he should not want to work at all. Some of the boys in the focus groups saw work as infringing on time for sports and video games, a narrative Ice seemed to support. Nate pushed back a bit, however, saying that a teen might want to work when they have done everything else. This provides us with useful data on views about work. For instance, we can contemplate how Ice's comments may be linked to the reproduction of a “slacker boy” narrative that we have addressed elsewhere (Sheppard et al., 2019), as well as Nate's disruption of this narrative. The example also shows us how these young people are attempting to shape each other's views of work for young people their age.
Gender dynamics also came up in some side conversations, perhaps related to attempts to shape each other's views but maybe also for some boys to position themselves as uninvested in the gendered work of babysitting (Sheppard et al., 2019). For example, when discussing school versus work, and whether to prioritize a baseball game or babysitting, some boys were quite dismissive and silly around the idea of babysitting. This happened in the main focus group conversations, but also in some side comments. For example, in a side comment about which should be prioritized, school or work, Frit (a boy) built on previous comments about the need to prioritize school by saying “you don’t want to be a babysitter your whole life,” which led another boy to laugh. In another example, a silly conversation shifted between the main group and side comments and was focused on what they should do with the children they were babysitting while also playing the baseball game. At one point one of the boys, Cameron, said in a side comment, “tie them on a leash.” Also, the above-discussed comment of “who really cares?” was in response to a discussion about meeting a babysitting commitment. Together, these side comments, alongside many of the comments from the main research discussion, reinforced a dismissive attitude towards babysitting, especially among many of the boys. With Allen (2005), we might see these interactions as an example where teens work through gendered identities in focus groups.
We also see girls participating in the gendered regulation of comments. For example, we see this in an exchange between Thad (a boy) and K (a girl), when discussing the scenario where a girl is being sexually harassed at work. Thad explained how he would deal with the scenario and then K responded:
[laughing] Thad, you’re gonna get touched?
No, I said “I’m out [leaving the job].”
Thad, you are not a girl [laughs].
I’m saying if I was.
In this exchange, K seemed to be questioning Thad's ability to speak on this topic because of his gender. While seemingly playful, K's comment can be considered an attempt to silence Thad because he is speaking as a boy, and perhaps she feels he does not adequately recognize the challenges of the scenario. Her comments are also problematically positioning Thad as a boy who therefore cannot imagine how he might react to sexual harassment. In another exchange there was a conversation about doing babysitting for a single parent. The facilitator asked, “What if it's a single mom?” at which point a boy, we think it was Ice, said “Then uhm [she should] get a boyfriend.” Berry picked up on how this is a troubling statement and, in a side comment, said “Oh my god.” A disagreement then developed between Ice and Berry about whether Rebecca had said “single mom” or “single parent.” In these examples, we see girls pushing back against boys’ gendered comments in a way that we would have missed if we had disregarded this data.
In general, we found that in the side comments the participants were more likely to be blunt and regulatory with each other than in the wider group. One example illustrated a heated side interaction during the discussion of the babysitting versus baseball scenario. Instead of having to choose, Jake optimistically suggested: “Your parents have to drive you to the baseball game. You could just ask if they can watch the kid while you play baseball.” Adrianna challenged Jake, asking why parents would be at the game. In response, Austin shouted to Adrianna “Excuse me, excuse me, is the … Put your hand down, it's my turn.” But Adrianna responded with “No!” In this interaction participants were jockeying to speak and challenging each other's ideas. We wondered if this was also a gendered dynamic, with Austin displaying a kind of masculine bravado and attempting to silence Adrianna, but Adrianna resisting being interrupted (Allen, 2005). In another, starker example, Thad made various cutting side comments to K. At one point, he said “K, we don’t need a lecture” and at another point, he interrupted K to say “Wait for your turn, K.”
In contrast, we see another example where a girl was encouraged to speak on the group's position that it is better not to work at their age. The following exchange began with the main discussion and then shifted into a side conversation between Nate, Ice, and Isabel:
… you wouldn't want to work right now?
No
Yeah, like we said …
C’mon, Isabel, I haven’t heard you talk.
We want to hear from your side of the table.
Yah, Isabel, why don’t you talk?
As we have noted, some researchers argue that focus groups allow for some participants to remain silent if they prefer, unlike a one-on-one interview where a participant is expected to speak (see Katainen and Heikkilä, 2020). As we see in this example, Isabel is being pressured by others to speak. This interaction can be read as supportive encouragement, as undermining a participant's desire to be quiet, and/or resistance to the research topics and process (Nairn et al., 2005). In this exchange, Isabel is the only one without work experience. Echoing arguments made by Katainen and Heikkilä (2020), previous or current work experience might have offered some participants knowledge that was advantageous in the context of the focus group. Lived experience with the focus group topic might have foster more confidence for some to participate.
As we analyzed these side conversations, we were also alert to any patterns that seemed linked to participants’ racial or class-based locations. We did note one pattern. In this section, almost all of the exchanges that we highlighted were between white boys and racialized girls. K's background is South Asian, Berry and Isabel are both Black, and Adrianna did not answer the question about race. K, Berry, and Adrianna all challenged some of the comments made by white boys. It may be that these girls were particularly determined to intervene with some of the boys’ comments and were less constrained by the white, middle class, “nice girl” imperative of emphasized femininity (Hill, 2019). However, we also see that K was overtly told “we don’t need another lecture,” pointing to the risks and bravery of speaking up. In contrast, some boys encouraged Isabel to talk when she might not have wanted to. This could be an example of Isabel being positively encouraged by her white peers, or they could have been criticizing her for not wanting to speak up.
This section reminds us that focus groups are a rich setting for observing group dynamics, including patterns around age and power relations linked to gender (Allen, 2005), that are well beyond the hierarchy between the researcher and the participants (Halkier, 2010; Katainen and Heikkilӓ, 2020; Kitzinger, 1994; Raby, 2010; Spyrou, 2018). Reflecting arguments made by Duggleby (2005) and Farnsworth and Boon (2010), we note that these dynamics can be strikingly evident in side comments that are not always recorded, recognized or examined in data analysis.
Conclusion
Our focus groups were busy, with lots of contributions and side conversations. The extent of these side conversations seemed to be facilitated by the location (a cafeteria where many of the participants normally socialize), timing (right before the holidays and school break), the make-up of the focus groups (peers and friends at school), and Rebecca's facilitation style. Sometimes the side conversations were frustrating for us because they made the focus groups difficult to facilitate and, later, to transcribe. These side conversations were frequently missed by us while the focus group was happening, were hard to follow, and at first seemed to suggest a lack of engagement with the research topic. However, our analysis suggests that these conversations were frequently related to either the research topic, thus providing insight into participants’ thoughts on early work, or the research process. Methodologically, our analysis provides insight into peer dynamics in focus groups and provokes reflection on facilitation.
Although a challenge, transcribing the side conversations has helped us reflect on focus group facilitation and analysis, especially with young people. Engaging with the side conversations helped us to learn about the participants’ views of early work, to see how meanings about work were produced in these groups, and to get a deeper sense of the peer relations at work in this meaning-making. The side conversations revealed that our research questioning might not have been as clear as intended, and that it is important to pursue ways of better engaging with diverse young people about topics like work. For example, the side conversations where participants expressed confusion or even dissent with the questions we asked, provoked us to think more deeply about why our questions fell flat or were uninteresting or unrelatable for the participants (Allen, 2005; Katainen and Heikkilä, 2020; Raby, 2010), and how we might better prompt about their views on and experiences with early work.
In our turn to side conversations in focus groups, we have also noted ways that young people regulate and socialize each other in research contexts, for example, in terms of who speaks, who is invited or pressured to speak, who dissents, and who disrupts the focus groups conversation. Gendered peer dynamics shaped the regulation of talk and jockeying for participation, including whose ideas are validated, who takes up space in the focus group, and in expressing ideas around what is valuable work. These peer dynamics challenge a framing of a focus group facilitator as holding direct authority and control over the unfolding of the focus group. Further, we have illustrated how side conversations in focus groups provide insight into the ways that focus groups allow participants to make meaning together (Hollander, 2004; Warr, 2005). In some cases, these side conversations pointed to peer-to-peer education and miseducation about early work expectations and health and safety. Many of the side conversations we have discussed make clear that side conversations were spaces where participants leaned on each other to seek clarification about the questions asked, to gain reassurance about their ideas, to disagree, to encourage, and discourage each other. An implication of this finding is that focus group facilitators might find some value in opening up space in focus group for side comments to happen.
Engaging with side conversations does pose certain challenges, however. Sometimes it was difficult for us to determine where the main conversation ended, and a side conversation began. Further, groups where such side comments were more prominent were generally ones that were more loosely facilitated, leading to more difficult transcription. Finally, we have had to reflect on the ethics of working with side comments. Our research topic; the transparent, public set-up of our focus groups; and our anonymized reporting of these peer-to-peer exchanges led us to feel confident that it is ethical to look at these side conversations. However, side conversations raise concerns when addressing more personal topics, in more private settings where participants are not fully aware of digital recorders, and/or where the data may be revealing personal details. In these latter cases, member-checking would be a very useful addition.
Overall, side conversations that at first seemed to be overly chaotic and off-topic proved to be rich data. They gave us valuable insight into how our young participants interacted and made meaning together, and allowed us to hear from participants who were less vocal in the main group. Our focus on side comments and conversations in focus groups adds to existing scholarship about focus group participants’ silliness, silences, interruptions (Nairn et al., 2005), resistance and withdrawal (Jacobsson and Åkerström, 2012), identity work (Allen, 2005), and counter- (Heikkilä and Katainen, 2021) and cross-talk (Raby, 2010). Our engagement with data created through side comments and conversations illuminates the diversity and richness of qualitative data. Openness to messy and chaotic research encounters, which first might appear “off track” from the research agenda or even “failed data,” can lead to fruitful reflections on research topics, reveal participants’ questions and quibbles with the research and researchers, and offer insight into interaction dynamics between participants, especially when they are previously known to each other. We invite researchers to tune into this messy and (sometimes) chaotic nature of focus groups, and to thus rethink the boundaries of what counts as data in qualitative research. While the side conversations were at-times both “off track” and “on point” we feel that there is promise in thinking and writing about vibrant, lively data that might otherwise have been overlooked.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by Brock University Council for Research in the Social Sciences, Western University Faculty of Social Science, Brock University Social Justice Research Institute, Social Sciences and Humanities Research Council of Canada.
